The Rise of Mobile Photogrammetry: Mapping Our World with Smartphones
We all take photos. Constantly, often compulsively, thanks to the camera we carry with us in the form of a smartphone. An estimated five billion photographs are taken each day (PhotoAiD). But what happens when photography moves beyond personal memories and becomes a powerful tool for scientific monitoring and environmental analysis? The answer lies in a technique called photogrammetry, and it’s being revolutionized by the cameras in our pockets.
From Battlefield Reconnaissance to Environmental Science
Photography’s role in monitoring began in the early 20th century, during World War I. Aerial reconnaissance planes equipped with cameras were used to map enemy positions. This evolved during World War II, and after the conflict, the technology expanded into scientific applications. While aerial and satellite photography remained the primary methods, the last three decades have seen a convergence of photogrammetry with artificial intelligence and advanced computing, creating new possibilities.
What is Photogrammetry?
Photogrammetry is a technique used to extract measurements from objects – two or three-dimensional – by processing, measuring, and interpreting photographs. Though its origins trace back to the early 1900s, it truly matured in the 1990s with the development of digital photogrammetry software.
Why is Marine Photogrammetry So Popular?
Today, photogrammetry is a widely used technique in environmental monitoring, particularly in marine environments, for two key reasons:
- It allows for the reconstruction of 3D models of objects, enabling detailed analysis in the lab and reducing the demand for physical sampling, which can be challenging underwater.
- It requires relatively low budgets while still delivering high-quality results.
All that’s needed to get started is a camera and software for processing the photos, along with the necessary technical expertise.
From Field to Data: Setting Transects and Spatial References
Let’s break down the two key phases of photogrammetric technique: data acquisition and software processing.
Defining the Underwater Workspace: Metric Tape and Visibility
Before capturing images, it’s essential to define a “transect” – an area where data will be collected. This transect should be clearly marked, for example, with a light-colored metric tape underwater.
GPS Coordinates and Orientation: Why They Matter for 3D Models
This workspace must then be linked to real-world coordinates, allowing it to be located on a map. Typically, the starting point of the transect is recorded using GPS, ensuring accurate correspondence between the real and digital locations. This is crucial because the software reconstructs the 3D model based on the coordinates of the various points captured.
To further refine accuracy, a leveling bubble is fixed to the seabed, and its GPS coordinates and inclination are recorded to provide a better reference for the three-dimensional orientation of the monitoring site (Daniel T. I., et al. 2020).
Acquiring Images: Speed, Angle, and Overlap
With the setup complete, the image acquisition can initiate. Standard procedures recommend swimming at a constant speed around the subject, maintaining an angle as perpendicular as possible from above, at a distance of 1-2 meters.
Photogrammetry and Overlap: The 50–75% Rule
Images must overlap to allow the software to construct a unified image of the subject and its environment. Though, the overlap rate should be between 50% and 75% (Daniel T. I., et al. 2020), increasing to 80% in particularly murky water (NOAA, 2019).
Acquisition Errors: When Software Fails
Incorrect image acquisition can lead to software failures or inaccurate results. Careful attention to these details is paramount.
Software Processing: From Point Clouds to 3D Meshes
After acquiring the data, the software converts the photos or video frames into a 3D model of the subject. This is a complex process, especially when dealing with intricate objects like coral reefs or archaeological artifacts.
Selecting the Best Frames: Noise, Defects, and Discards
The first step is selecting the best photographs, discarding those with noise or defects that could compromise processing.
Key Points and “Structure from Motion”
The software then extracts key features from the images and reconstructs the angles from which they were taken. Using the entered spatial coordinates, it places the points in space, first creating a “point cloud” and then a 3D mesh – a collection of polygons (triangles or quadrilaterals) that form the three-dimensional surface of the object (Erica Nocerino, et al. 2020).
Cleaning and Refining: The Importance of Post-Processing
The model is then cleaned and refined through a process called “post-processing.” The final model can then be 3D printed or shared with other researchers for study without the need to physically remove or visit the site (Erica Nocerino, et al. 2020).
Future Trends in Mobile Photogrammetry
The accessibility of high-quality smartphone cameras is driving innovation in photogrammetry. Here’s what People can expect to spot in the coming years:
- AI-Powered Automation: Expect more automated workflows, with AI handling tasks like image alignment, point cloud generation, and mesh creation.
- Real-Time 3D Reconstruction: Advances in processing power will enable real-time 3D reconstruction on mobile devices, allowing for immediate visualization and analysis in the field.
- Integration with AR/VR: Photogrammetry data will be seamlessly integrated with augmented and virtual reality applications, creating immersive experiences for environmental monitoring and education.
- Citizen Science Initiatives: Mobile photogrammetry will empower citizen scientists to contribute to environmental monitoring efforts, collecting valuable data on a large scale.
- Underwater Drone Integration: Combining photogrammetry with underwater drones will allow for mapping and monitoring of even larger and more inaccessible marine environments.
FAQ
Q: What is the minimum camera resolution needed for photogrammetry?
A: While higher resolution is generally better, modern smartphone cameras with 12 megapixels or more can produce acceptable results.
Q: What software is commonly used for photogrammetry?
A: Popular options include Agisoft Metashape, RealityCapture, and Meshroom.
Q: Is photogrammetry expensive?
A: The cost can vary, but using a smartphone and free or low-cost software makes it a relatively affordable technique.
Q: Can photogrammetry be used for objects other than marine environments?
A: Absolutely! It’s used in archaeology, architecture, engineering, and many other fields.
Did you know? Smartphones now capture 92.5% of all pictures taken, leaving only 7.5% to conventional cameras (PhotoAiD).
Pro Tip: Ensure consistent lighting and avoid shadows when capturing images for photogrammetry. This will improve the accuracy of the 3D model.
What are your thoughts on the future of mobile photogrammetry? Share your ideas in the comments below!
